Recycled PET Streams Explained: Post-Consumer, Post-Industrial and Closed Loop

Polyethylene terephthalate (PET) is a linear aromatic polyester produced by melt-phase esterification of purified terephthalic acid with ethylene glycol, followed by vacuum polycondensation. The intrinsic viscosity of PET, measured by ASTM D4603-18 or ISO 1628-5, typically ranges from 0.55 dL/g for fibre grades to 0.84 dL/g for bottle grades. Virgin PET intended for food-contact rigid packaging is manufactured to comply with 21 CFR 177.1630. Recycled PET streams are distinguished not primarily by polymer chemistry but by contamination profile, thermal history, residence time and regulatory pathway. The three principal classifications are post-consumer, post-industrial and closed loop. A post-consumer flake with an intrinsic viscosity of 0.74 dL/g may be unacceptable for bottle-to-bottle production without decontamination, while a post-industrial regrind with similar intrinsic viscosity may be readily re-extruded under controlled drying conditions. The table below summarises typical stream characteristics reported in industrial extrusion and moulding operations.

Post-Consumer rPET Feedstock Heterogeneity and Contamination Boundaries

Post-consumer recycled PET enters the extrusion feed system as washed hot-washed flake from municipal deposit-return or single-stream collection. Bale composition varies by source: deposit-return bales commonly contain 95–98 wt% PET bottle flake, while curbside mixed plastic bales may deliver 70–85 wt% PET with polyolefin caps, labels, silicone adhesives and occasional polyvinyl chloride fractions. The critical contaminant is PVC, which degrades below PET processing temperatures and releases hydrogen chloride. Production-scale melt filtration configured with 100–200 µm screen changers and hydrocyclone density separation is required when PVC content exceeds 50 ppm. Residual label adhesives and sink/float carryover introduce ash levels up to 0.2 wt% unless hot-wash caustic concentration and residence time are maintained. The intrinsic viscosity of washed flake typically falls from virgin bottle-grade 0.82 dL/g to 0.70–0.78 dL/g because hydrolytic chain scission occurs during washing and extrusion. To compensate, processors either blend 20–40 wt% post-consumer flake with virgin resin or route flake through solid-state polymerization at 210–220°C under vacuum. Extrusion melt temperature is usually capped at 285°C; above this threshold acetaldehyde generation accelerates and yellowness index increases. Final pellets are tested according to ASTM D4603-18 for intrinsic viscosity and ASTM D3418 for melting and crystallization enthalpies.

Why Does Post-Industrial Regrind Demand Tighter Drying Control?

Because post-industrial PET is generated from sprues, runners, edge trim, off-spec preforms, purge lumps and start-up scrap on manufacturing floors before consumer use, the molecular weight distribution has already been subjected to one or more melt histories. The main processing risk is not particulate contamination but chain extension variability and residual moisture in regrind. Extruders processing 100% post-industrial regrind often operate with twin-screw venting and vacuum pump sets that maintain barrel zone vacuum below 30 mbar absolute. The feed section must be purged with dried air when ambient relative humidity exceeds 60%; otherwise moisture absorbed on flake surfaces hydrolyses ester linkages and reduces intrinsic viscosity by 0.02–0.05 dL/g within a single pass. Crystallized post-industrial flake is routinely dried at 160–170°C for 4–6 h in desiccant bed dryers to reach residual moisture below 50 ppm, but production-scale hopper residence times shorter than 3 h leave moisture gradients that produce inconsistent melt pressure. Melt pressure variation exceeding ±1.5 MPa across a screen changer indicates either insufficient drying or feed bridging. In injection moulding, clamp force settings for PET preform tools are often specified at 4–6 kN/cm² of projected area. Post-industrial regrind containing mixed copolymer grades can widen the melt viscosity envelope enough to cause short shots or gate blush when nozzle temperature is below 270°C. Melt flow rate measured at 280°C under 2.16 kg load according to ISO 1133-1:2022 is used to compare lot-to-lot variation, although intrinsic viscosity remains the primary specification.

In sheet extrusion operations producing thermoformed food trays from post-industrial trim, the absence of a dedicated heading does not reduce the need for raw-material traceability. Edge trim generated from extruded PET sheet at 0.5–1.5 mm thickness retains orientation stresses and often contains recycled coextruded barrier tie layers or antifog additives that shift crystallization half-times. When such trim is reintroduced at 30–50 wt% into the main feed, extruder screw speed and barrel zone 2 temperature may require adjustment because regrind bulk density is 0.45–0.60 g/cm³ versus 0.80–0.85 g/cm³ for virgin pellets. The feeder must therefore operate in gravimetric mode rather than volumetric mode to hold throughput variation below ±2%. Vacuum venting at 20–40 mbar absolute is needed to remove ethylene glycol formed by hydrolytic chain scission. Without sufficient venting, residual glycol acts as an internal plasticizer and lowers melt strength, causing web sag and draw resonance at calendar stack speeds above 25 m/min. Sheet producers commonly use a screen changer with 200–300 µm mesh and a gear pump to maintain die head pressure above 100 bar. Post-industrial PET re-utilized within the same manufacturing site may fall under general reuse provisions, but when scrap crosses company boundaries it becomes a commercial recycling stream subject to supplier audit and incoming intrinsic viscosity verification using ISO 1628-5.

When Closed-Loop Bottle-to-Bottle Streams Require Solid-State Repolymerisation

Closed-loop PET recovery converts used bottles into food-contact pellets that re-enter bottle production without downgrading into fibre or strapping. This stream is distinguished by decontamination performance rather than source purity alone. A closed-loop process must reduce migration of surrogate contaminants such as toluene, chlorobenzene, phenylcyclohexane and benzophenone to below 10 µg/kg in food simulants under challenge test conditions; European Food Safety Authority evaluations and FDA letters of no objection reference these surrogates. Because post-consumer bottle flake retains residual monomers, acetaldehyde and low-molecular-weight oligomers, solid-state repolymerisation is used after melt extrusion and pelletization. SSP reactors operate at 190–215°C under vacuum below 1.5 mbar with nitrogen purge for 10–16 h; this raises intrinsic viscosity from 0.72 dL/g to 0.82–0.84 dL/g and simultaneously removes residual acetaldehyde to below 2 ppm. The process window is narrow: if SSP temperature exceeds 218°C, amorphous pellets sinter and block the reactor bed; if temperature falls below 190°C, ester-interchange and polycondensation become too slow to achieve intrinsic viscosity targets within a standard 14 h residence. Pre-drying before SSP is required at 170°C for 5 h to bring pellet moisture below 30 ppm. Extruder melt temperature during closed-loop pellet production is typically limited to 280°C because higher temperature accelerates formation of vinyl ester end groups and acetaldehyde. In preform injection, dried closed-loop rPET is processed at 270–285°C with barrel residence time under 5 min; screw L/D ratio for PET is generally 20:1–24:1 with low shear compression to limit viscous heating. If preform intrinsic viscosity falls below 0.78 dL/g after moulding, bottle top-load strength decreases and environmental stress crack resistance in carbonated beverage packaging becomes insufficient.

Analytical verification across the three streams requires separation of recycling route from polymer test result. A post-industrial pellet with intrinsic viscosity of 0.80 dL/g can still fail food-contact suitability if the source was non-food-contact packaging. Conversely, a post-consumer flake with intrinsic viscosity of 0.74 dL/g may pass EFSA challenge testing after super-clean recycling. Quality assurance therefore couples rheological measurements with thermal desorption GC-MS, headspace GC for acetaldehyde, and migration testing under food simulant conditions. Sorting infrastructure also determines yield: near-infrared units using spectral ranges around 4.7–10.5 µm reject polylactic acid and PVC, optical colour sorters remove opaque fractions, and metal detectors protect extruder screws from ferrous and non-ferrous tramp metal. Incoming flakes that fail a sink/float separation test at specific gravity 1.32–1.38 g/cm³ are diverted because PET density is 1.38–1.40 g/cm³ in amorphous form and 1.45–1.50 g/cm³ in crystalline form. Published data for long-term mechanical property retention of multi-loop closed-loop rPET beyond five recycling cycles is limited, and industrial practice therefore favours blending 25–50 wt% super-clean rPET with virgin PET to maintain cumulative intrinsic viscosity loss below 0.05 dL/g. The operational boundary is defined by contaminant accumulation, not by polymer chain length alone.